Hi-

For a long time I've thought that projects like SETI@home were pretty cool.
Recently, the thought worked its way back to the forefront of my brain and I went to my favorite source for more information about distributed computing.
I signed up my home computer to work on projects for the World Community Grid.
Anyone have some thoughts or experiences to share?

WCG allows 'teams' to sum their points (for no real benefit other than looking like fun statistics). There isn't an empeg team... yet!

Apparently Folding@home has had some good results. I would get in on some of these programs, but my computer hasn't been very reliable, and the software that most of these programs distributes hasn't been very good, IMO.

My only problem is that since it makes your CPU run at 100% all the time, it ends up costing a bit in power comsumption, both in electricity for processing, and in heat generated by the CPU that then has to be air conditioned. Not major issues, but I don't like my chip staying hot when I'm not using it.

I used to be really big into distributed.net and was ranked pretty high. One big reason that I stopped though was due to moving to more thermal controlled systems with dynamic fans. First my Shuttle XPC, and most recently to laptops. Having to hear what would normally be a quiet machine turn into a beast as noisy as full desktop PCs got on my nerves quickly.

I also see no concrete benefits from these programs. Many sound good in theory, but lack any real world results now after years of being out there. I'd much rather donate my money directly to the scientific communities involved in real research instead of handing that money over to the local energy company by having my system run 24/7 full tilt.

I run this 'distributed computing' screensaver:

http://www.electricsheep.org/

The results are clearly visible, pretty pictures to save your screen

I've been running WCG for maybe 9 months now on numerous computers. I have no idea if it will benefit the biomedical community, but I sure hope it will.

I don't see how a computer running 100% or running 2% (idle) are any different on power input. My power supply is still pulling in the same amount regardless. WCG or not, I never shut my computer off and never put it in any stand by or power save (because no computer ever seems to reliably recover from those).

The WCG Windows client (United Devices) is simple for casual users and causes minimal slowdown to the system. The cross platform client (Boinc) allows for running many different computing projects, allows remote management, allows multi processor and more. And there is an active community on the WCG forums to help with configuration, troubleshooting, deployment/management automation, etc.

During the summer, the CPU heat was a problem since some of my computers were in un-airconditioned spaces. Remotely, I was able to have them run only at night. Later, the ability was built in to throttle CPU usage to any percentage.

I am glad to use 100% of my CPU time toward a project which may have some potential. If I'm not using it, then it's wasted.

That Electric Sheep project looks very very cool. The family tree is very interesting, seeing how influence of two parent images can be seen in a child image.

Windows Media Player visualizations, eat your heart out.

I know you're loath to upgrade, but surely you're not still running a CPU that doesn't have an idle state. Modern CPUs (where "modern" probably means "engineered in the last decade") consume much less power when they're not doing anything than when they're running full bore. That's why fans kick on: to dissipate the additional heat generated by the additional power consumption.

Personally, I'm fine with trading that off with the computer doing something, but don't fool yourself into thinking you're getting something for nothing.

Just because your power supply is CAPABLE of pulling 350 watts, doesn't mean it necessarily IS.

Good points, all around.

My Athlon C 1.4 likely does have an idle state, but the fans don't know about it and run full even at idle. The newer machines I manage are much better about this.

I've got some electrical engineering to learn before I'll truely know if my "enegry conservation efforts" are paying off. Too many assumptions (always turn off CRTs and room lights, but leave flourescents on because they consume more power when starting up) and not enough information (I thought the PSU always drew 350 watt).

What I do know is, my "energy conservation" is nullified by so many inefficient and wasteful systems out there. And my energy waste is similarly dwarfed by them. Makes me wonder what is the power consumption of the entire WCG (or any other distro computing project)? Maybe we should opt for having a super computer do it slower but with less total (project life) power consumption.

Actually, a flourescent bulb will draw about 2.5 times it's running current to start up, so a 15 watt light will draw 45 watts, for about 2 seconds. Even giving it 5 seconds, starting up the light will only cost you 15 seconds of running current.

The stuff in these lights will probably die faster by running hours rather that amount of starts.

Yeah but does it even matter when the bulbs last so long. I had one compact flourescent that lasted five years.

I don't think it matters, but it was a point raised in the previous post.

I've had a couple die in a year, but they were the cheap 99 cent ones. Which are all I have.

http://www.cnn.com/2006/TECH/fun.games/09/18/playstation.folding/index.html

Seems the PS3 will be folding when it is idle. I wonder how many units will be bought just to plug into some server room as a cheep alternative to a bunch of Cell powered servers.